Model card briefing
DeepSeek-V4.1-Flash-1017 is a minor update to DeepSeek-V4.1-Flash delivering substantial improvements across reasoning, coding, agentic tasks, and visual benchmarks. Alongside those gains, it reduces task execution time and output length, enabling faster completion at lower cost and advancing the performance–cost Pareto frontier across multiple agent harnesses.
Extended README benchmark comparison
| Benchmark | DS-V4.1-Flash-1017 | DS-V4.1-Flash | K3 | GLM-5.3 | GPT-5.6 Sol | Opus 5 |
|---|---|---|---|---|---|---|
| HLE w/o tools | 42.6 (44.0†) | 36.8 (39.1†) | 43.5 | 42.0† | 44.5 | 56.3 |
| Terminal-Bench 4.0 | 42.3 | 31.2 | 12.6 | 37.9 | 39.9 | 51.8 |
| DeepSWE v1.1 | 77.0 | 74.2 | 67.5 | 66.9 | 73.0 | 74.0 |
| ProgramBench | 29.1 | 20.3 | 17.5 | 19.0 | 23.0 | 37.0 |
| NL2Repo-Bench | 68.2 | 65.4 | 58.0 | 58.0 | 56.8 | 75.3 |
| Automation-Bench v1.0.6 | 60.5 | 54.8 | 46.7 | 48.8 | 45.8 | 50.3 |
| Agents’ Last Exam | 35.4 | 31.8 | 27.6 | 28.5 | 26.7 | 28.6 |
| Apex Bench v1.1 | 72.1 | 64.6 | 47.8 | 56.6 | 54.3 | 65.8 |
| Chartography (w/tools) | 83.0 | 78.9 | 68.1 | — | 79.9 | 84.0 |
| BabyVision (w/tools) | 92.5 | 89.6 | 85.7 | — | 88.9 | 94.1 |
| ZeroBench-main (w/tools) | 51.0 | 49.0 | 41.0 | — | 53.0 | 52.0 |
† denotes the text-only subset of HLE. The publisher uses the Claude Code harness for Terminal-Bench 4.0 and the Mini-SWE-V2 harness for DeepSWE v1.1.
Prompt encoding
This base release does not include a Jinja-format chat template. The repository’s encoding folder contains a self-contained Python reference implementation with test cases for multi-turn conversations, tool calling, thinking mode, numeric reasoning effort, mid-conversation system messages, and interleaved image content.
For production use, the publisher also releases deepseek-recipe: Rust libraries with Python bindings that convert Messages, Chat Completions, and Responses API requests into the Conversation format, encode DeepSeek V4/V4.1 prompts or token IDs, and parse complete or streamed responses. Model inference, tool execution, and HTTP transport remain with the caller.
Minimal inference and sampling
The README points to its inference folder for weight conversion and local inference instructions.
| Publisher parameter | Recommended value |
|---|---|
| temperature | 1.0 |
| top_p | 1.0 |
| context_window | 1M tokens |
| max_tokens | ≥256K |
Citation: DeepSeek-AI, “DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression,” 2026.